1. Field of the Invention
The present invention relates to a recombinant antibody composition which is a human IgG1 antibody, comprises a CH2 domain in which amino acids at positions 276 and 339 indicated by the EU index as in Kabat, et al. are replaced by other amino acids and has more improved complement-dependent cytotoxic activity than an antibody comprising a CH2 domain before the amino acids are replaced; a DNA encoding the antibody molecule or a heavy chain constant region of the antibody molecule contained in the recombinant antibody composition; a transformant obtainable by introducing the DNA into a host cell; a process for producing the recombinant antibody composition using the transformant; and a medicament comprising the recombinant antibody composition as an active ingredient.
2. Brief Description of the Background Art
Since antibodies are protein molecules having high binding activity and binding specificity to a target molecule (antigen) and high stability in blood, applications thereof to diagnostic, preventive and therapeutic agents for various human diseases have been attempted (Non-patent Document 1). Although antibodies are generally produced by administering (immunizing) an antigen to a non-human animals antibodies obtained from a non-human animal have an amino acid sequence specific to the species and side effects are caused due to that the antibodies are recognized as foreign substances in the human body. Accordingly, human chimeric antibodies or humanized antibodies have been prepared from antibodies of animals other than human (nonhuman animals) using gene recombination techniques (Non-patent Documents 2 to 5).
The human chimeric antibodies and humanized antibodies have resolved problems possessed by non-human animal antibodies such as mouse antibodies, such as the high immunogenicity, low effector function and short blood half-life, and applications of monoclonal antibodies to pharmaceutical preparations were made possible by using them (Non-patent Documents 6 to 9). In the Unites States, for example, a plurality of humanized antibodies have already been approved as an antibody for cancer treatment, and are on the market (Non-patent Document 10).
These human chimeric antibodies and humanized antibodies actually show effects to a certain degree at clinical level, but therapeutic antibodies having higher effects are in demand. For example, in the case of single administration of RITUXAN™ (Non-patent Document 11) (manufactured by IDEC/Roche/Genentech) which is a human chimeric antibody to CD20, it has been reported that its response ratio for recurrent low malignancy non-Hodgkin lymphoma patients in the phase III clinical test is no more than 48% (complete remission 6%, partial remission 42%), and its average duration of response is 12 months (Non-patent Document 12). In the case of combination use of RITUXAN™ and chemotherapy (CHOP: Cyclophosphamide, Doxorubicin, Vincristine), it has been reported that its response ratio for recurrent low malignancy and follicular non-Hodgkin lymphoma patients by the phase II clinical test is 95% (complete remission 55%, partial remission 45%), but side effects due to CHOP were found (Non-patent Document 13). In the case of single administration of HERCEPTIN™ (manufactured by Genentech) which is a humanized antibody to HER2, it has been reported that its response ratio for metastatic breast cancer patients in the phase III clinical test is only 15%, and its average duration of response is 9.1 months (Non-patent Document 14).
The human antibody molecule is also called immunoglobulin (hereinafter referred to as Ig) and classified into isotypes of IgA1, IgA2, IgD, IgE, IgG1, IgG2, IgG3, IgG4 and IgM based on its molecular structures IgG1, IgG2, IgG3 and IgG4 having relatively high homology in amino acid sequences are genetically called IgG. Human IgG is mainly used as a therapeutic antibody.
An antibody molecule comprises two kinds of polypeptides, i.e. a heavy chain (hereinafter referred to as H chain) and a light chain (hereinafter referred to as L chain). A human IgG antibody molecule comprises two H chains and two L chains. Also, an H chain comprises an H chain variable region (hereinafter referred to as VH) and an H chain constant region (hereinafter referred to as CH), and an L chain comprises an L chain variable region (hereinafter referred to as VL) and an L chain constant region (hereinafter referred to as CL). The H chain constant region comprises four domains which are respectively called CH1, hinge, CH2 and CH3 domains from the domain close to VH located at the heavy chain N-terminal in this order. Also, the CH2 domain and CH3 domain in combination are called Fc.
An antibody binds to an antigen via an antigen-biding region (hereinafter referred to as Fv) comprising VH and VL and binds to an effector molecule in the immune system such as a receptor or a complement via the H chain constant region. Under the mediation of the binding to the effector molecule in the immune system, the antibody induces an effector activity such as a complement-dependent cell-mediated cytotoxic activity (hereinafter referred to as CDC activity), an antibody-dependent cellular cytotoxic activity (hereinafter referred to as ADCC activity) or a phagocytic activity so as to eliminate the antigen or cells (a pathogen or tumor cells) expressing the antigen.
To induce the ADCC activity or phagocytic activity, it is important that the antibody binds to a member of the Fc gamma receptor (hereinafter referred to as FcγR) family expressed on the surface of various leukocytes such as natural killer cells (hereinafter referred to as NK cells), monocytes, macrophages or granulocytes. The FcγR family includes activated FcγR and regulated FcγR. FcγRI, FcγRIIa, FcγRIIIa and FcγRIIIb belong to the activated FcγR, and FcγRIIb belongs to the regulated FcγR. A human IgG antibody strongly binds to such a receptor and consequently induces the ADCC activity or phagocytic activity of leukocytes.
The ADCC activity is a reaction in which leukocytes such as NK cells mainly lyse targets cell under the mediation of an antibody. The antibody binds to an antigen on the surface of the target cells via Fv and binds to FcγRIIIa on the surface of NK cells via Fc. As a result, the NK cells release cytotoxic molecules such as perforin or granzyme and thus lyse the target cells (Non-patent Documents 15 and 16).
The CDC activity is a reaction in which a group of serum proteins called complements lyses target cells under the mediation of an antibody. The complements are classified into C1 to C9 proteins, and they are subjected to chain reaction to thereby induce the CDC activity. Each of the complement proteins is activated by reacting with a specific complement protein and then reacts with the subsequent complement protein. These chain reactions start with the binding of the first complement component C1 to the Fc of an antibody, which has been bonded via Fv to an antigen on the surface of target cells, via C1q that is one of the proteins constituting C1. Finally, complexes of C5 to C9 are polymerized together to form a hole in the cell membrane of the target cells, which results in the lysis of the target cells (on-patent Documents 15 and 16).
Four human IgG isotypes (IgG1, IgG2, IgG3 and IgG4) are highly homologous with each other in the amino acid sequence in the H chain constant region except for the hinges showing a wide variety. However, these isotypes induce an effector activity of different strengths (Non-patent Document 17). In general, the ADCC activity decreases in the following order: IgG1>IgG3>IgG4≧IgG2 (Non-patent Documents 18 and 19), while the CDC activity decreases in the following order: IgG3≧IgG1>>IgG2≈IgG4. As discussed above, the binding of an antibody to C1q is important in inducing the CDC activity. The biding constants (Ka) in the binding of C1q to a monomeric antibody molecule in human IgG isotypes, i.e., IgG1, IgG2, IgG3 and IgG4 are 1.2×104, 0.64×104, 2.9×104 and 0.44×104, respectively (Non-patent Document 20), reflecting the difference in CDC activity among these isotypes.
Concerning the drug effect mechanisms of clinically employed antibody drugs, the importance of ADCC and CDC activities has particularly attracted public attention. It is reported that RITUXAN™ as described above, which is a human chimeric antibody of the IgG1 isotype, shows ADCC and CDC activities in vitro (Non-patent Document 21). Relating to the clinical effects of RITUXAN™, it is reported that RITUXAN™ shows high therapeutic effects on a patient showing a genotype with high ADCC activity (Non-patent Document 22), that complement components in blood are quickly consumed following the administration thereof (Non-patent Document 23), that the expression of CD59, which is a CDC activity regulator, shows an increase in cancer cells of a patient suffering from recurrence after the administration thereof (Non-patent Document 24), and the like. These reports indicate that RITUXAN™ actually exerts the effector function in the body of a patient. It is also reported that HERCEPTIN™ as described above, which is a humanized antibody of the IgG1 subclass, shows the ADCC activity in vitro (Non-patent Document 25).
Although human IgG1 and human IgG3 are isotypes having excellent ADCC and CDC activities, it is known that human IgG3 antibody has a shorter half life in the blood than other human IgG isotypes and thus quickly disappears from the blood after the administration (Non-patent Document 26). It is also known that human IgG3 has no protein A-binding activity, differing from other human IgG isotypes (Non-patent Document 27). In producing an antibody on an industrial scale, a purification process using protein A is predominant and other processes using, for example, protein G have some problems such as a high purification cost.
It is known that protein A binds to a human IgG antibody molecule (Non-patent Document 28). When indicated by the EU index as in Kabat, et al. (Non-patent Document 29), it is pointed out as the results of X-ray crystallographic analysis that a loop comprising the amino acids at positions 252 to 254, a loop consisting of the amino acids at positions 308 to 312, and a loop comprising the amino acids at positions 433 to 436 are important (Non-patent Document 28). As the results of nuclear magnetic resonance (NMR) analysis, it is further indicated that Ile253, Ser254, His310, Gln311, His433, His435 and His436 are particularly important in the Fc of IgG1 (Non-patent Document 30). Furthermore, Kim, et al. found that the protein A-binding activity was attenuated by replacing His435 of a human IgG1 with Arg435 derived from IgG3 (Non-patent Document 31). Hereinafter, the positions of the amino acids in the amino acid sequence of an antibody molecule are represented based on the EU index as in Kabat, et al. (Non-patent Document 29).
Based on the above it can be said that human IgG1 antibody is the most suitable isotype as an antibody drug, since it has higher ADCC and CDC activities than other isotypes, can be purified using protein A, shows a long half life in blood and has a merit from the viewpoint of production cost. Although a human IgG1 antibody has been employed as drugs in practice as described above, the drug effects exhibited by the existing antibody drugs are still insufficient. Thus, there has been required an antibody drug having improved effects. In order to satisfy this requirement, studies have been made on an antibody having enhanced effector activities. As discussed above, an effector activity of an antibody reflects the binding activity of the H chain constant region to an effector molecule in the immune system. Accordingly, the effector activity of the antibody can be enhanced by enhancing the binding activity of the H chain constant region to the effector molecule in the immune system.
In order to analyze the effector activities of human antibodies, studies have been made on antibodies comprising two kinds of human isotype amino acid sequences which are prepared by partly swapping the amino acid sequences in the heavy chain constant region between two kinds of human isotype antibodies having different effector activity (Patent Document 1 and Non-patent Documents 32 and 33). In late 1980's, Morrison, et al. indicated that antibody molecules, which were prepared by swapping the individual domains (CH1, CH2, CH3 and hinge) in the heavy chain constant region between IgG1 having a high effector activity and IgG4 having a low effector activity, or between IgG2 having a low effector activity and IgG3 having a high effector activity, could be expressed as recombinant proteins (Patent Document 1). As the results of the subsequent analysis on these antibody molecules, they have clarified that the C-terminal side of the CH2 domain is important in the CDC activity of IgG1 and the CH2 domain is important in the CDC activity of IgG3 (Non-patent Document 32); the CH2 domain and hinge are important in the binding of IgG1 and IgG3 to FcγRI (Non-patent Document 33); and the like.
As described above, the CH2 domain is important in the CDC activity. The amino acid sequences of human IgG1 antibody and human IgG3 antibody having high CDC activity have been analyzed. Concerning the amino acid sequences of CH2, it is known that Leu235 (Non-patent Document 34), Asp270, Lys322, Pro329 and Pro331 (Non-patent Document 35) are important in the CDC activity of human IgG1; and Gly233, Leu234, Leu235, Gly236 (Non-patent Document 36) and Lys322 (Non-patent Document 37) are important in the CDC activity of human IgG3. Brekke, et al. analyzed various antibody molecules prepared by transplanting amino acid residues being common to the CH2 domain amino acid sequences of human IgG1 antibody and human IgG3 antibody having high CDC activity or several amino acid residues being different from a human IgG4 antibody having very low CDC activity into a human IgG4 antibody. As a result, they found that the CDC activity of human IgG4 antibody was enhanced by swapping Ser331 in human IgG4 by Pro331 which is common to a human IgG1 and a human IgG3 (Non-patent Document 38).
Moreover, attempts have been made to enhance the CDC activity by swapping a part of the amino acid sequence of the heavy chain constant region of human IgG3 antibody, which is the human IgG isotype having the highest CDC activity, by an amino acid sequence originating in another human IgG isotype. Concerning the hinge lengths of each IgG isotypes, IgG1 has 15 amino acid residues, IgG2 has 12 amino acid residues, IgG3 has 62 amino acid residues and IgG4 has 12 amino acid residues. Thus, the human IgG has a structural characteristic of having a longer hinge than other IgG3 isotypes Non-patent Document 1). The hinge of human IgG3 antibody consisting of 62 amino acids is encoded by four exons on a gene. Michaelsen, et al. reported that the CDC activity of human IgG3 antibody having a hinge that was shortened to 15 amino acid residues by deleting three exons in the N-terminal side among these four exons was higher than IgG3 and IgG1 (Non-patent Document 39). Norderhang, et al. reported that the CDC activity is further increased by swapping the amino acid sequences of the hinge shortened in the above and the amino acid sequences of the hinge of IgG4. Further, Brekke, et al. reported that when the hinge of human IgG3 antibody was swapped by the hinge of human IgG1 antibody, the CDC activity of the resultant antibody was higher than IgG3 and similar to IgG1 or more (Non-patent Document 41).
On the other hand, studies have been made on an antibody prepared by replacing the amino acid sequence of the heavy chain constant region of human IgG1 antibody by an artificial amino acid sequence which is not present in the nature to thereby increase the C1q-binding activity and thus enhance the CDC activity (Non-patent Document 42 and Patent Documents 2 to 5). As described above, the CDC activity is induced by the binding of C1q, which is one of the proteins constituting complement protein C1, to the Fc of an antibody molecule. Idusogie, et al. reported that by replacing Lys326 or Glu333 in the CH2 domain of RITUXAN™ (a human IgG1 chimeric antibody) as described above with an other amino acid, the CDC activity was enhanced twice at most (Non-patent Document 42, Patent Document 2). Furthermore, Idusogie, et al. indicated that by replacing Lys326 or Glu333 in IgG2 with an other amino acid, the CDC activity of IgG2, which inherently corresponds to a several hundredth part of the CDC activity of IgG1, was increased to about one over twenty-five of IgG1 (Patent Documents 3 to 5).
However, such an antibody prepared through the replacement of an amino acid sequence which is not present in the nature has a risk that it is recognized as a foreign matter in the human body and thus induces a side effect similar to the non-human animal antibody as discussed above. On the other hand, the amino acid sequence of an antibody prepared by swapping amino acid sequences between human isotypes is a combination of amino acid sequences of antibodies inherently carried by humans.
In the therapeutic effects of a therapeutic antibody, the ADCC and phagocytic activities induced by the biding of the Fe region of the antibody to FcγR and the CDC activity mediated by the biding of the antibody to C1q are both important. However, the bindings of the antibody to C1q and to the FcγR are both mediated by the Fc and, therefore, it is feared that an amino acid modification aiming to enhance the CDC activity might damage the ADCC activity. In practice, Idusogie, et al. reported that an antibody in which the CDC activity was enhanced by replacing the Fe of human IgG1 antibody with an artificial amino acid sequence showed a serious lowering in the ADCC activity (Non-patent Document 42).
As a procedure for enhancing an effector activity of an antibody other than the replacement in an amino acid sequence, regulation of a sugar chain attached to the constant region of the antibody may be cited. It is known that the ADCC activity of human IgG antibody changes based on the structure of a complex-type N-glycoside-linked sugar chain attached asparagine at position 297 in the Fc (FIG. 1 shows a model view thereof) (Patent Document 6). It is also reported that the ADCC activity of the antibody changes depending on the amounts of galactose and N-acetylglucosamine contained in this sugar chain (Non-patent Documents 43 to 46). However, the ADCC activity is mostly affected by fucose binding to N-acetylglucosamine in the reducing terminal through α1,6-bond in the sugar chain. Namely, an IgG antibody having complex-type N-glycoside-linked sugar chains in which fucose is not bound to N-acetylglucosamine in the reducing terminal in the sugar chains shows remarkably higher ADCC and FcγRIIIa-binding activities than an IgG antibody having complex-type N-glycoside-linked sugar chain in which fucose is bound to N-acetylglucosamine in the reducing terminal in the sugar chains (Non-patent Documents 47, 48 and 49 and Patent Document 7). Although antibody molecules having no fucose in sugar chains exist in vivo as a natural-type, α1,6-fucosyltransferase gene-knockout cells have been known as cells capable of specifically producing an antibody composition having complex-type N-glycoside-linked sugar chains in which fucose is not bound to N-acetylglucosamine in the reducing terminal in the sugar chains (Patent Documents 7 and 8)    Non-patent Document 1: Monoclonal Antibodies: Principles and Applications, Wiley-Liss, Inc. (1995)    Non-patent Document 2: Nature, 312, 643 (1984)    Non-patent Document 3: Proc. Natl. Acad. Sci. USA, 81, 6851 (1984)    Non-patent Document 4: Nature, 321, 522 (1986)    Non-patent Document 5: Nature, 332, 323 (1988)    Non-patent Document 6: Immunol. Today, 21, 364 (2000)    Non-patent Document 7: Immunol. Today, 21, 403 (2000)    Non-patent Document 8: Ann. Allergy Asthma Immunol., 81, 105 (1998)    Non-patent Document 9: Nature Biotechnol., 16, 1015 (1998)    Non-patent Document 10: Nature Reviews Cancer, 1, 119 (2001)    Non-patent Document 11: Curr. Opin. Oncol., 10, 548 (1998)    Non-patent Document 12: J. Clin. Oncol., 16, 2825 (1998)    Non-patent Document 13: J. Clin. Oncol., 17, 268 (1999)    Non-patent Document 14: J. Clin. Oncol., 17, 2639 (1999)    Non-patent Document 15: Chemical Immunology, 65, 88 (1997)    Non-patent Document 16: Immunol. Today, 20, 576 (1999)    Non-patent Document 17: Monoclonal Antibodies: Principles and Applications, Wiley-Liss, Inc. (1995)    Non-patent Document 18: Nature, 332, 323 (1988)    Non-patent Document 19: Journal of Experimental Medicine, 166, 1351 (1987)    Non-patent Document 20: Biochemistry, 15, 5175 (1976)    Non-patent Document 21: Oncogene, 22, 7359 (2003)    Non-patent Document 22: Blood, 29, 754 (2002)    Non-patent Document 23: J. Immunol., 172, 3280 (2004)    Non-patent Document 24: J. Clin. Oncol, 21, 1466 (2003)    Non-patent Document 25: Cancer Immunol Immunother., 37, 255 (1993)    Non-patent Document 26: Cancer Res., 58, 3905 (1998)    Non-patent Document 27: Scand. J. Immunol., 15, 275 (1982)    Non-patent Document 28: Biochemistry, 20, 2361 (1981)    Non-patent Document 29: Sequence of Proteins of Immunological Interest, Fifth Edition (1991)    Non-patent Document 30: FEBS Lett., 328, 49 (1993)    Non-patent Document 31: Eur. J. Immunol., 29, 2819 (1999)    Non-patent Document 32: Journal of Experimental Medicine, 173, 1025 (1991)    Non-patent Document 33: Journal of Experimental Medicine, 173, 1483 (1991)    Non-patent Document 34: Immunology, 86, 319 (1995)    Non-patent Document 35: J. Immunol., 164, 4178 (2000)    Non-patent Document 36: Mol. Immunol., 34, 1019 (1997)    Non-patent Document 37: Mol. Immunol., 37, 995 (2000)    Non-patent Document 38: Ear. J. Immunol., 24, 2542 (1994)    Non-patent Document 39: Scand. J. Immunol., 32, 517 (1990)    Non-patent Document 40: Eur. J. Immunol., 21, 2379 (1991)    Non-patent Document 41: Mol. Immunol., 30, 1419 (1993)    Non-patent Document 42: J. Immunol., 166, 2571 (2001)    Non-patent Document 43: Human Antib Hybrid, 5, 143 (1994)    Non-patent Document 44: Hum Antib Hybrid, 6, 82 (1995)    Non-patent Document 45: Nat. Biotechnol, 17, 176 (1999)    Non-patent Document 46: Biotechnol Bioeng., 74, 288 (2001)    Non-patent Document 47: Clin. Cancer. Res., 10, 6248 (2004)    Non-patent Document 48: J. Biol. Chem., 277, 26733 (2002)    Non-patent Document 49: J. Biol. Chem., 278, 3466 (2003)    Patent Document 1: US2003/0158389A1    Patent Document 2: WO00/42072    Patent Document 3: US2004/0132101 A1    Patent Document 4: US2005/0054832 A1    Patent Document 5: WO00/61739    Patent Document 6: WO02/31140    Patent Document 7: WO03/85107